This guide explains how to use the Brand Safety Guidelines Template for setting up and customizing brand safety preferences tailored to a client’s requirements. The tracker is a vital tool for communicating a brand's unique needs to the managed services team, ensuring that the social media content is audited according to the client's specific guidelines.
Configuring Brand Safety Guidelines (BDGs)
The first step is defining the brand's tolerance levels for over 20 safety violation categories. For detailed explanations of each category, refer to this help article.
Turning Categories ON/OFF
The ON/OFF toggle in the first column is used to record brand's high level decisioning on that violation category. Note the toggle does not directly control the platform's filtering system, however it is used by the auditing team to prioritize violation categories and make decisions at the post level if needed.
When a category is turned ON, content that AI flags as violation will be sent to the human auditing team for a manual review. The auditing team will then assess the content based on provided guidelines and reject any post they deem a threat to the client’s brand's safety standards.
Choosing "Any" vs. "Excessive"
For the Alcohol, Cannabis, and Profanity categories, there are two levels of sensitivity may be selected:
- Any: This setting indicates a zero tolerance policy by the client. Any instance of the violation will be flagged for review.
- Excessive: This setting triggers a more strict review of content that features repeated, glamorized, or prolonged instances of the violation.
Note: For these three categories, this is a binary either/or scenario, both scenarios cannot be selected.
Using the "Notes" for Client Nuances
Every flagged post is reviewed by the auditing team, specific instructions or exceptions can be provided in the “NOTES” column. This allows for nuanced enforcement that matches the brand's preferences.
For example, the following notes can be added:
- Arms & Ammunition: "Accept in gaming context".
- Profanity: "Accept in music".
- Sexual or nude content: "Accept in artwork".
Adding clear notes ensures the auditing team applies the direction consistently and provides an additional level of customization for each brand.
Tracking Keywords and Phrases
The Tracked Keywords & Phrases section provides an additional layer of brand safety by detecting specific words the client would like reviewed. This feature is designed to function as its own violation category, where a human review will distinguish the keyword's context and tone.
Important: Keywords that are already covered by the main violation categories should not be added to the keyword list. For example, there is no need to add "Guns" as a keyword, as any post containing it would already be flagged under the "Arms & Ammunition" category.
Examples of Keywords used by brands:
| Socio-Political | "Abortion," "elections," specific political party names. |
| Industry-Specific | Competitor names, executive names, or specific products. |
| Legal | "DUI," "DWI", "lawsuit," "litigation." |
Keyword Modes: Exact Match
- Exact Match: This mode finds only results that are a perfect letter-for-letter match to the specified keyword. For instance, a search for "apple" will only detect the word "apple."
For further clarification, feel free to reach out to your dedicated Product Success Manager.